Backing up to alternate external USB 2.0 80 GB HDD 1

[coyote1]

I want to backup Ghost images of several PC's to two alternate ext. USB 80 GB hard drives weekly (one drive offsite for the alternating week) accessable across the network using the technigue as described in the following article:

http://www.pcstats.com/articleview.cfm?articleid=1676&page=9.

I would like to have both USB drives to be able to share the same permantent mounted folder name.
Right now I have to create two unique named folders each representing the individual USB drives which complicates matters. Maybe I am making this more complicated that it is ?

Appreciate any suggestions,

Karl

 

[wahnula]

Hmmm...usually, it's the drive letters, not the names that cause issues. But, while connected to main machine, go to My Computer, right-click the drive in question, and select 'rename' then call it whatever. Mine are "USB1" and USB2" but in your case you would want them the same.

One thought...Ghost is a decent backup tool, but I believe to recover you need the EXACT same hardware as the original machine. As machines age it gets harder and harder to find the same chipset, etc. If the original rig is stolen or burns to the ground (my worst fear) you will be tasked to assemble the exact same chipset PC to recover.

One solution (that I use) is to have the hardware-specific backup drive as a cheap resident IDE drive, for nightly imaging, and pair this backup with a multi-drive file-copy backup like SyncBackSE which saves everything in the format it was created. Browsing these drives is like browsing the original folder. Since one remains offsite, after a disaster I can get my company working again by connecting the drive to a slave computer, mapping the drive, and everyone can get to their files. I also backup the Exchange Store to these drives, but that will take an Exchange Server to read, still the files are there.

Just some thoughts on my obsession, backing up. My diligence paid off recently when my SBS RAID 1 OS array went poo-poo, I replaced the drives and recovered from the resident backup drive (also quicker then USB) and was back up with no loss in hours.

Tony

 

[coyote1]

Tony,

Very creative !

You have me thinking. I am going to re-configure things tonight and let you know what my scenario turns out to be.
You are right about the chipsets but if worse comes to worse I was hoping to be able mount the image and browse to restore the specific file or folders to a different PC if need be ?

Thanks for your time,
Karl

 

[Freestone]

Acronis' Universal Restore addition to their True Image product line has enabled me to restore images to very dissimilar hardware with very goods results. Granted their server line of products is pricey, but the workstation versions of these products aren't too badly priced. Something to consider looking into if disaster recovery is on your agenda.

 

[coyote1]

Thanks for the tip

 

[wahnula]

Disaster recovery should be a multi-pronged approach combining as many methods as you can realistically support.

http://www.mozy.com

has a decent online backup program that's affordable, but reliable? We don't know yet, but it's there.

I learned early on that all the backups in the world won't help if they don't work when you need them to. Early in the life of the machine, I recommend simulating a disaster (or actually creating one, like removing a HDD). Put your plan into action and see how robust it is, how long it takes, good & bad points, etc.

Windows has a good backup built in called NTbackup, it works well as does their ASR, another prong in the approach.

When I build a mission-critical machine, I ALWAYS use RAID 1 as a minimum, sometimes a combo of RAID 1 and RAID 5, this gives some built in resiliency for hardware failure. But the real bugaboo (in my case) is fire and theft, and my butt is only as safe as my latest backup. People spend thousands on backup schemes, only to let them languish until it's too late...they have a problem and only a 3-month old backup.

Tape, hard drives, flash, remote, I don't care, just DO IT and then remember to TEST IT so you know it will work. There is no better feeling than bringing a machine back form the dead, and going from panicked to relaxed and even smug.

Of course nobody will ever appreciate this but you.

Tony

 

[coyote1]

Hi Tony,

I concure with what you say. What I am thinking of doing for this office (small budget using existing parts and software) is to have both workstations creating a weekly scheduled Ghost image to a network share on the server's spare internal HDD. I then will Xcopy or NTbackup the server's local C: drive and spare drive (w/other workstation images) to an external USB 2.0 80GB IDE drive which will alternate weekly offsite.
In the future I want to setup a software mirrored drive pair on the server for availability of common data stores in the event of a drive failure during a business day.

Thanks for your input,

Karl

 

[wahnula]

In the future I want to setup a software mirrored drive pair

Karl,

Be careful here. Software RAID is not as resilient as hardware RAID, plus if you use PCI/PCIe/PCI-X hardware RAID controllers (I like 3ware) you can change server hardware with autonomy.

On your backup plan: why are your workstations backing up to the server? It is more typical to have the workstations file's reside on the server precluding the need for individual backup. If it's an HDD size limitation, 500GB HDDs are less than $100 now, drop one of those in your server. Then ALL the workstation's files are already on the server, with the exception of email and calendar files.

I know your pain, I started with a few workstations, one running MS Calendar, then hired a company that sold me an uber-expensive (dual P-Pro NT 4.0 Exchange BackOffice 5.5 Veritas tape backup) server until its 10 GB SCSI drive was full. I said screw them and installed SBS 2003 OEM myself onn some new hardware. Picked up the software (legally) for around $200, it will run on ANYTHING (very low min. reqs). I would recommend this to you as you go forward, as now will you not only have all files stored on the server you will also have all the email and calendar residing there too. This is very handy if a workstation goes down, just swap it with any other. I also know the scrimp & scrounge method of putting a network together. I showed the boss the bill from the old server + software now I do whatever I want (always trying to save thought)!

On your backup plan it seems you will be backing up the workstations to a static network share then taking a clone of C: offsite; it should be the other way around. Backup your C: and whatever else is needed to make your server run (basically anything you have a disk for) plus whatever files will fit on the spare local drive, then take home the "work product" which I imagine would be the workstation's files on a two-or-three disk rotation. Hope I'm not confusing you too much!

Tony

 

[coyote1]

Tony,

The workstations are customized for very specialized purposes (have to have access 24/7) and thus the need to able to re-create the environment in a short period of time by using drive images. We had to do this already when one of the workstation's drives would not boot and we had a deadline!). All digital assets and data have a common data store on the main workstation/server through standardized mapped drive/shares. Right now I am do a scheduled cloning of the workstation/server C drive and the other two workstation C drive clone/images to this spare server internal drive(D drive) then run a scheduled after hour xcopy batch file to copy all three clone/images to the external USB drive which rotates offsite.

I origonally did all this to the spare rotating removable D drive off the server/workstation but had to take the system down when swapping the drive (pain in the butt). With this new configuration I am hoping to avoid this scenario.

Next year I want to replace these systems and build a fault tolerant system that will implement some of the ideas you have expressed.

I am allways looking for a way to make things simpler and to be able to recover in a short period of time so I appreciate your input and interest.

Thanks,

Karl

 

[wahnula]

Just shows to go ya...there are as many backup scenarios as there are usage scenarios. It sounds like your plan is well thought out for your application, taking your unique needs into account. Many pros see my backup plan as overkill and way too much work but I know it will work...for me...and that is the most important thing. YOUR comfort level is what matters, as well as your confidence in your disaster recovery. It sounds like you had a test of the backup system not too long ago and it passed, good for you. So many folks do a backup routine that they have no idea whether it will work or not if needed...a pet peeve of mine.

One caveat about those USB drives...if it's a Windows system, make sure they are stopped using "Safely Remove Hardware" before removing them. It's a simple thing but can corrupt an entire drive if you remove one that's in the process of being written to. If you are unable to Stop the drive, try logging off & back on, that usually does it. My next backup system will feature eSATA drives for just this reason.

Tony

 

[coyote1]

Tony,

I am currently not "Safely Remove Hardware" before removing them in regards to the USB drives. I will take your reccomendation since that corruption issue is something I definitely want to avoid !!

Great tip thanks,


Karl